Learning about sustainable adoption of green agricultural technologies: Experimental evidence from training with organic farming practices and soil tests in Indonesia
Economic Policy, Applied Economics
Final Report Abstract
The agricultural sector faces numerous environmental challenges, some of which are caused by unsustainable agricultural practices such as the overuse of fertilizers and pesticides. This project used an experimental design to evaluate the long-term impact of organic farming training in Indonesia, a country struggling with land degradation and overuse of agricultural inputs, but where policymakers and non-governmental organizations also show strong interest in sustainable farming. We built upon a previous project that investigated the short-term impact of organic farming training. The previous project employed a randomized controlled trial with 1,200 smallholder farmers and three waves of data collection (2018, 2019, and 2020). Half of these farmers were randomly selected to receive organic farming training in 2019, while the other half, the control group, received no training. In this project, we added a second round of training in 2022. The training involved the same farmers as in 2019, while the control group remained unchanged. We further implemented a fourth wave of data collection in 2023. This approach created a unique dataset on smallholder farming practices and adoption behavior spanning six years, four data waves, and two training interventions. The randomization of training access enabled us to examine the causal effect of training, avoiding biases related to farmers’ characteristics and external factors. The relevance of the project’s findings is twofold: First, this project contributes to the limited research differentiating between initial and longer term adoption. While short-term effects can be informative, understanding the true impact of information provision requires a better understanding of the longer-term consequences. Moreover, research is still scarce regarding the effect of repeated information provision. Our findings indicate that after two rounds of training, farmers invited to the training are significantly more likely to use organic inputs. Most farmers who adopt organic farming practices combine them with chemical inputs, which they partially reduce. Although farmers do not fully adopt organic farming, the reduction in nitrogen application is significant given the high rates of chemical fertilizer use in Indonesia. Our quantitative and qualitative data show that the adoption process of organic farming practices is complex and non-linear. These findings can inform policymakers in Indonesia and similar contexts when designing extension programs. Second, evidence on the causal effect of adopting organic farming (practices) remains limited. Many existing studies do not control for possible selection bias. Yet, if more talented, richer, or more motivated farmers self-select into organic farming, a comparison between organic and conventional farmers is misleading. In contrast, the randomized design of this project allows us to analyze the causal effect with respect to adoption and welfare. Our results indicate that adopting organic farming practices does not strongly affect welfare indicators measured in the quantitative data. Concerns that increased adoption of organic farming practices and reductions of chemical inputs negatively impact yields or increase labor costs were not supported. Additionally, qualitative data suggest that farmers perceive the training as empowering as it reduces their dependence on chemical fertilizer subsidies, which have been decreasing in recent years.
